Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Ahh cool, thx. :) Am 19. Oktober 2017 13:03:04 MESZ schrieb Patricia Shanahan : >In general, we are documenting our release process in >https://cwiki.apache.org/confluence/display/OOOUSERS/How+to+Cook+a+Release. > >We have a problem that the ASF itself documents the rules for voting on > >releases in two places: >http://www.apache.org/legal/release-policy.html#release-approval and >https://www.apache.org/foundation/voting.html > >Those rules do not quite work for us. In practice, we treat a justified > >-1 from any community member as a veto. We put much more emphasis on >testing binaries than on building from source. I suppose in theory when > >we see a justified -1 all PMC members should change their votes to -1, >and change them back to +1 if and when the problem is resolved. > >On 10/19/2017 12:06 AM, Peter kovacs wrote: >> Have we documented our process somewhere. I am unsure. I think we >have at least discussed this already. >> All arguments sound so familiar. >> >> Am 19. Oktober 2017 02:35:29 MESZ schrieb Dave Fisher >: >>> On Oct 18, 2017, at 4:45 PM, Pedro Lino >>> wrote: > Maybe we need to ask for review of >>> http://www.apache.org/legal/release-policy.html#release-approval at >the >>> same time as looking at the voting process documentation. If taken >>> literally, a PMC member who cannot do builds from source can't cast >a >>> +1 vote, because their vote is binding and a binding +1 requires a >>> build from source. Probably the ASF wants to guarantee that at least 3 PMC members are >>> developers (or development inclined)? >>> >>> It’s not probably. It *IS*! >>> > What I would like is to change it to require at least three PMC >>> members to declare they have done a build from source and tested the >>> result. Other PMC members could vote based on binary testing and >>> signature checking without building. Actually for a multi-platform software such as AOO it should be >>> required that building from source and testing the result was >carried >>> out by at least one voter (PMC member or not) for each of the >>> platforms/bit depths. If all three PMC members have success in building on e.g. Linux x64 >>> it does not provide any guarantee for the other platforms (as proven >by >>> 4.1.4 RC4) >>> >>> We had two PMC providing the community builds. The official release >is >>> the Source release. We need as much testing as possible of the >>> community releases. AOO is a unique project for Apache because our >>> users count on the community builds and not the source releases. >>> >>> I think we have grown in the last year because in the first years >here >>> at Apache most of the knowledge on how to build was in the minds of >the >>> former Hamburg team - Star, Sun, Oracle and then IBM employees. >>> >>> Special thanks to Matthias and Jim how providing the Community >Builds. >>> >>> Regards, >>> Dave >>> >- To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org >> >> - >> To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org >> For additional commands, e-mail: dev-h...@openoffice.apache.org >> > >- >To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org >For additional commands, e-mail: dev-h...@openoffice.apache.org - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
I would like to go a step further and go for auto builds with no manual involvement. For this we must be able to produce same binaries independent from the build machine so we can verify the build by beeing able to create a reference build. In order to ensure that what we build is the result from the source code we use. And there we will have to look that we get 3 reference builders that verify our binary. Am 19. Oktober 2017 13:32:28 MESZ schrieb Patricia Shanahan : >On 10/18/2017 5:35 PM, Dave Fisher wrote: >> >>> On Oct 18, 2017, at 4:45 PM, Pedro Lino >wrote: >>> >>> >>> Maybe we need to ask for review of >http://www.apache.org/legal/release-policy.html#release-approval at the >same time as looking at the voting process documentation. If taken >literally, a PMC member who cannot do builds from source can't cast a >+1 vote, because their vote is binding and a binding +1 requires a >build from source. >>> >>> Probably the ASF wants to guarantee that at least 3 PMC members are >developers (or development inclined)? >> >> It’s not probably. It *IS*! > >If that is the purpose, it is not a very efficient way of achieving it. >It would take less of my time, for example, to attach an abbreviated CV >that I could write once and use for each vote, perhaps just a list of >degrees, patents, and former employers. > >> >>> What I would like is to change it to require at least three PMC >members to declare they have done a build from source and tested the >result. Other PMC members could vote based on binary testing and >signature checking without building. >>> >>> Actually for a multi-platform software such as AOO it should be >required that building from source and testing the result was carried >out by at least one voter (PMC member or not) for each of the >platforms/bit depths. >>> If all three PMC members have success in building on e.g. Linux x64 >it does not provide any guarantee for the other platforms (as proven by >4.1.4 RC4) >> >> We had two PMC providing the community builds. The official release >is the Source release. We need as much testing as possible of the >community releases. AOO is a unique project for Apache because our >users count on the community builds and not the source releases. >> >> I think we have grown in the last year because in the first years >here at Apache most of the knowledge on how to build was in the minds >of the former Hamburg team - Star, Sun, Oracle and then IBM employees. >> >> Special thanks to Matthias and Jim how providing the Community >Builds. > >I really get convinced that the source code is good for building >purposes by seeing people I trust upload binaries for each >combination of language and supported platform. I then download a few >of >the binaries and test them. > >During the vote period I do a ritual build of one combination on my own >hardware, and test the resulting binary, just to satisfy the ASF rules. >I would rather spend more time testing binaries that people are going >to >use rather than testing a binary I'm going to throw away. > >- >To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org >For additional commands, e-mail: dev-h...@openoffice.apache.org - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
On 10/18/2017 5:35 PM, Dave Fisher wrote: On Oct 18, 2017, at 4:45 PM, Pedro Lino wrote: Maybe we need to ask for review of http://www.apache.org/legal/release-policy.html#release-approval at the same time as looking at the voting process documentation. If taken literally, a PMC member who cannot do builds from source can't cast a +1 vote, because their vote is binding and a binding +1 requires a build from source. Probably the ASF wants to guarantee that at least 3 PMC members are developers (or development inclined)? It’s not probably. It *IS*! If that is the purpose, it is not a very efficient way of achieving it. It would take less of my time, for example, to attach an abbreviated CV that I could write once and use for each vote, perhaps just a list of degrees, patents, and former employers. What I would like is to change it to require at least three PMC members to declare they have done a build from source and tested the result. Other PMC members could vote based on binary testing and signature checking without building. Actually for a multi-platform software such as AOO it should be required that building from source and testing the result was carried out by at least one voter (PMC member or not) for each of the platforms/bit depths. If all three PMC members have success in building on e.g. Linux x64 it does not provide any guarantee for the other platforms (as proven by 4.1.4 RC4) We had two PMC providing the community builds. The official release is the Source release. We need as much testing as possible of the community releases. AOO is a unique project for Apache because our users count on the community builds and not the source releases. I think we have grown in the last year because in the first years here at Apache most of the knowledge on how to build was in the minds of the former Hamburg team - Star, Sun, Oracle and then IBM employees. Special thanks to Matthias and Jim how providing the Community Builds. I really get convinced that the source code is good for building purposes by seeing people I trust upload binaries for each combination of language and supported platform. I then download a few of the binaries and test them. During the vote period I do a ritual build of one combination on my own hardware, and test the resulting binary, just to satisfy the ASF rules. I would rather spend more time testing binaries that people are going to use rather than testing a binary I'm going to throw away. - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
In general, we are documenting our release process in https://cwiki.apache.org/confluence/display/OOOUSERS/How+to+Cook+a+Release. We have a problem that the ASF itself documents the rules for voting on releases in two places: http://www.apache.org/legal/release-policy.html#release-approval and https://www.apache.org/foundation/voting.html Those rules do not quite work for us. In practice, we treat a justified -1 from any community member as a veto. We put much more emphasis on testing binaries than on building from source. I suppose in theory when we see a justified -1 all PMC members should change their votes to -1, and change them back to +1 if and when the problem is resolved. On 10/19/2017 12:06 AM, Peter kovacs wrote: Have we documented our process somewhere. I am unsure. I think we have at least discussed this already. All arguments sound so familiar. Am 19. Oktober 2017 02:35:29 MESZ schrieb Dave Fisher : On Oct 18, 2017, at 4:45 PM, Pedro Lino wrote: Maybe we need to ask for review of http://www.apache.org/legal/release-policy.html#release-approval at the same time as looking at the voting process documentation. If taken literally, a PMC member who cannot do builds from source can't cast a +1 vote, because their vote is binding and a binding +1 requires a build from source. Probably the ASF wants to guarantee that at least 3 PMC members are developers (or development inclined)? It’s not probably. It *IS*! What I would like is to change it to require at least three PMC members to declare they have done a build from source and tested the result. Other PMC members could vote based on binary testing and signature checking without building. Actually for a multi-platform software such as AOO it should be required that building from source and testing the result was carried out by at least one voter (PMC member or not) for each of the platforms/bit depths. If all three PMC members have success in building on e.g. Linux x64 it does not provide any guarantee for the other platforms (as proven by 4.1.4 RC4) We had two PMC providing the community builds. The official release is the Source release. We need as much testing as possible of the community releases. AOO is a unique project for Apache because our users count on the community builds and not the source releases. I think we have grown in the last year because in the first years here at Apache most of the knowledge on how to build was in the minds of the former Hamburg team - Star, Sun, Oracle and then IBM employees. Special thanks to Matthias and Jim how providing the Community Builds. Regards, Dave - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Have we documented our process somewhere. I am unsure. I think we have at least discussed this already. All arguments sound so familiar. Am 19. Oktober 2017 02:35:29 MESZ schrieb Dave Fisher : > >> On Oct 18, 2017, at 4:45 PM, Pedro Lino >wrote: >> >> >> >>> Maybe we need to ask for review of >http://www.apache.org/legal/release-policy.html#release-approval at the >same time as looking at the voting process documentation. If taken >literally, a PMC member who cannot do builds from source can't cast a >+1 vote, because their vote is binding and a binding +1 requires a >build from source. >> >> Probably the ASF wants to guarantee that at least 3 PMC members are >developers (or development inclined)? > >It’s not probably. It *IS*! > >> >>> What I would like is to change it to require at least three PMC >members to declare they have done a build from source and tested the >result. Other PMC members could vote based on binary testing and >signature checking without building. >> >> Actually for a multi-platform software such as AOO it should be >required that building from source and testing the result was carried >out by at least one voter (PMC member or not) for each of the >platforms/bit depths. >> If all three PMC members have success in building on e.g. Linux x64 >it does not provide any guarantee for the other platforms (as proven by >4.1.4 RC4) > >We had two PMC providing the community builds. The official release is >the Source release. We need as much testing as possible of the >community releases. AOO is a unique project for Apache because our >users count on the community builds and not the source releases. > >I think we have grown in the last year because in the first years here >at Apache most of the knowledge on how to build was in the minds of the >former Hamburg team - Star, Sun, Oracle and then IBM employees. > >Special thanks to Matthias and Jim how providing the Community Builds. > >Regards, >Dave > >> >> - >> To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org >> For additional commands, e-mail: dev-h...@openoffice.apache.org >> - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
> On Oct 18, 2017, at 4:45 PM, Pedro Lino wrote: > > > >> Maybe we need to ask for review of >> http://www.apache.org/legal/release-policy.html#release-approval at the same >> time as looking at the voting process documentation. If taken literally, a >> PMC member who cannot do builds from source can't cast a +1 vote, because >> their vote is binding and a binding +1 requires a build from source. > > Probably the ASF wants to guarantee that at least 3 PMC members are > developers (or development inclined)? It’s not probably. It *IS*! > >> What I would like is to change it to require at least three PMC members to >> declare they have done a build from source and tested the result. Other PMC >> members could vote based on binary testing and signature checking without >> building. > > Actually for a multi-platform software such as AOO it should be required that > building from source and testing the result was carried out by at least one > voter (PMC member or not) for each of the platforms/bit depths. > If all three PMC members have success in building on e.g. Linux x64 it does > not provide any guarantee for the other platforms (as proven by 4.1.4 RC4) We had two PMC providing the community builds. The official release is the Source release. We need as much testing as possible of the community releases. AOO is a unique project for Apache because our users count on the community builds and not the source releases. I think we have grown in the last year because in the first years here at Apache most of the knowledge on how to build was in the minds of the former Hamburg team - Star, Sun, Oracle and then IBM employees. Special thanks to Matthias and Jim how providing the Community Builds. Regards, Dave > > - > To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org > For additional commands, e-mail: dev-h...@openoffice.apache.org > signature.asc Description: Message signed with OpenPGP
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Maybe we need to ask for review of http://www.apache.org/legal/release-policy.html#release-approval at the same time as looking at the voting process documentation. If taken literally, a PMC member who cannot do builds from source can't cast a +1 vote, because their vote is binding and a binding +1 requires a build from source. Probably the ASF wants to guarantee that at least 3 PMC members are developers (or development inclined)? What I would like is to change it to require at least three PMC members to declare they have done a build from source and tested the result. Other PMC members could vote based on binary testing and signature checking without building. Actually for a multi-platform software such as AOO it should be required that building from source and testing the result was carried out by at least one voter (PMC member or not) for each of the platforms/bit depths. If all three PMC members have success in building on e.g. Linux x64 it does not provide any guarantee for the other platforms (as proven by 4.1.4 RC4) - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
On 10/18/2017 07:18 PM, Gavin McDonald wrote: On 19 Oct 2017, at 9:39 am, Carl Marcum wrote: Question.. If a PMC member cast a vote should it always be counted as binding whether stated explicitly or not? Yes, Gav… Thanks Gavin !! - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
On 10/18/2017 07:21 PM, Andrea Pescetti wrote: Carl Marcum wrote: If a PMC member cast a vote should it always be counted as binding whether stated explicitly or not? Yes, it should. People add "binding" as a shorthand to mean "I am from the PMC" since a release cannot be approved without reaching 3 votes from the PMC, and it must be easy to verify this requirement. But once this requirement is satisfied, and assuming there aren't any -1 votes, adding "binding" or not really makes no difference, and I've omitted it at times. There is no such thing as "a non-binding vote from a PMC member". And even in the tally it is not really important to separate the two, provided one names 3 PMC members who voted +1 and points out that no -1 votes were cast. Regards, Andrea. Thanks for the clarification Andrea. - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
On 10/18/2017 07:16 PM, Patricia Shanahan wrote: On 10/18/2017 3:39 PM, Carl Marcum wrote: Question.. If a PMC member cast a vote should it always be counted as binding whether stated explicitly or not? The vote counting is really only for ASF rule conformance. The way it really works for AOO, even a single properly justified -1, regardless of source, stops the release in its tracks. Similarly, any +1 increases confidence in the release. Maybe we need to ask for review of http://www.apache.org/legal/release-policy.html#release-approval at the same time as looking at the voting process documentation. If taken literally, a PMC member who cannot do builds from source can't cast a +1 vote, because their vote is binding and a binding +1 requires a build from source. What I would like is to change it to require at least three PMC members to declare they have done a build from source and tested the result. Other PMC members could vote based on binary testing and signature checking without building. If we tried to follow the ASF policies literally only the handful of us who are set up to build AOO would have any voice at all in releases, which is ridiculous. That was kind of the way I was looking at it. It might be good to include the links in any documentation for the release vote email. Thanks, Carl - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Carl Marcum wrote: If a PMC member cast a vote should it always be counted as binding whether stated explicitly or not? Yes, it should. People add "binding" as a shorthand to mean "I am from the PMC" since a release cannot be approved without reaching 3 votes from the PMC, and it must be easy to verify this requirement. But once this requirement is satisfied, and assuming there aren't any -1 votes, adding "binding" or not really makes no difference, and I've omitted it at times. There is no such thing as "a non-binding vote from a PMC member". And even in the tally it is not really important to separate the two, provided one names 3 PMC members who voted +1 and points out that no -1 votes were cast. Regards, Andrea. - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
> On 19 Oct 2017, at 9:39 am, Carl Marcum wrote: > > > Question.. > If a PMC member cast a vote should it always be counted as binding whether > stated explicitly or not? Yes, Gav… > - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
On 10/18/2017 3:39 PM, Carl Marcum wrote: On 10/18/2017 03:11 PM, Jim Jagielski wrote: We expect PMC members to have a deep sense of responsibility to their projects. When they cast a binding vote, we trust that they have satisfied themselves enough to warrant them casting the vote. That, alone, should be sufficient, imo. - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org Thanks to Pedro for pointing out that language and to Dave for taking action on it. For the record I only counted binding votes when they were explicitly cast as such. Perhaps this is my error. I didn't base it on steps taken as listing them is not a requirement that I'm aware of. Question.. If a PMC member cast a vote should it always be counted as binding whether stated explicitly or not? The vote counting is really only for ASF rule conformance. The way it really works for AOO, even a single properly justified -1, regardless of source, stops the release in its tracks. Similarly, any +1 increases confidence in the release. Maybe we need to ask for review of http://www.apache.org/legal/release-policy.html#release-approval at the same time as looking at the voting process documentation. If taken literally, a PMC member who cannot do builds from source can't cast a +1 vote, because their vote is binding and a binding +1 requires a build from source. What I would like is to change it to require at least three PMC members to declare they have done a build from source and tested the result. Other PMC members could vote based on binary testing and signature checking without building. If we tried to follow the ASF policies literally only the handful of us who are set up to build AOO would have any voice at all in releases, which is ridiculous. - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
On 10/18/2017 03:11 PM, Jim Jagielski wrote: We expect PMC members to have a deep sense of responsibility to their projects. When they cast a binding vote, we trust that they have satisfied themselves enough to warrant them casting the vote. That, alone, should be sufficient, imo. - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org Thanks to Pedro for pointing out that language and to Dave for taking action on it. For the record I only counted binding votes when they were explicitly cast as such. Perhaps this is my error. I didn't base it on steps taken as listing them is not a requirement that I'm aware of. Question.. If a PMC member cast a vote should it always be counted as binding whether stated explicitly or not? Thanks, Carl - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Hi Dave, all I proposed the following new logic on the comdev mailing list. Who is permitted to vote is, to some extent, a community-specific thing. However, the basic rule is that only PMC members have binding votes, and all other votes are advisory. Much better with exactly the same meaning. It turns out that this awful language on the voting page goes back to at least July 2004. Thanks to Pedro for pointing it out. It should be changed in the next few days. Thank _you_ for listening. I still don't understand why ASF is/was so strict. ASF Software is still "distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied" (all the SHOUTING copied ipsis verbis from the ASF page http://www.apache.org/legal/src-headers.html#headers) So what is all the "rules are there to protect all PMC members with an insurance and a legal shield" argumentation about? The software is provided as is without warranties... Kind regards, Pedro
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
So the practice is much saner than the theory and nobody discarded Larry's -1 saying it was not coming from a PMC member. Any correctly justified -1 will of course be treated in the same way, regardless of who casts it. I'm aware of that. That is exactly why I applauded AOO's PMC members attitude just reading the ASF rules discouraged me to become a committer (I was invited in the past) and still does I did a presentation at ApacheCon a few years ago, called "Bending the Rules: Community over Code over Policy", explaining that policy should never come first and presenting some examples where we had to, say, "interpret" a dumb policy to do the right thing that would not hamper the community. It's still buried at https://community.apache.org/speakers/slides.html and that is still my point of view. But also something that concretely happens in OpenOffice. Actually I had already read your presentation. As I mentioned in a previous email I believe the community currently working on AOO is healthier and wiser than in the Oracle period. I'm just not sure about the "Apache way"... Seems like a very strict organization which is does not have the right mindset for a software that should IMO be for the benefit of the users (not for developers to have fun ;) ) Agreed. That is why the "*and all others are either discouraged from voting (to keep the noise down)*" is particularly unnecessary (if not offensive). I trust Dave will get that unfortunate wording amended in https://www.apache.org/foundation/voting.html - but again, that is about technical voting (on a code commit) and not release voting. In this sense, the sentence is a bit easier to justify (i.e., if you are expressing an opinion on a controversial code commit you are supposed to have familiarity with the code). That was just an example... Actually the wording in the Release voting is even worse. What is all the SHOUTING for? I'm glad all these rules are there to protect all PMC members with an "insurance and a legal shield"! And I'm glad you guys volunteered to be a PMC because this is the opposite of having fun ;) Best regards, Pedro - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
> On Oct 18, 2017, at 3:12 PM, Patricia Shanahan wrote: > > On 10/18/2017 12:28 PM, Pedro Lino wrote: >> On 18/10/2017 19:02, Andrea Pescetti wrote: >>> This is OK as long as the Apache Policy doesn't get in the way too much... >>> I mean, I highly appreciate when people provide a list of what they did to >>> justify their +1. But this shouldn't be used "against" them. >> +1 (as a user) >> ASF does a hell of a job to discourage participation! I particularly like >> this gem "the basic rule is that only PMC members have binding votes, *and >> all others are either discouraged from voting (to keep the noise down)* or >> else have their votes considered of an indicative or advisory nature only. >> I'm glad AOO PMC members take these recommendations with a pinch of salt. > > I particularly dislike the "only" in "advisory nature only". The success > of AOO depends on the advice of every community member who cares enough > to download a release candidate and test it. I proposed the following new logic on the comdev mailing list. Who is permitted to vote is, to some extent, a community-specific thing. However, the basic rule is that only PMC members have binding votes, and all other votes are advisory. It turns out that this awful language on the voting page goes back to at least July 2004. Thanks to Pedro for pointing it out. It should be changed in the next few days. Regards, Dave > > - > To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org > For additional commands, e-mail: dev-h...@openoffice.apache.org > signature.asc Description: Message signed with OpenPGP
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
On 10/18/2017 12:28 PM, Pedro Lino wrote: On 18/10/2017 19:02, Andrea Pescetti wrote: This is OK as long as the Apache Policy doesn't get in the way too much... I mean, I highly appreciate when people provide a list of what they did to justify their +1. But this shouldn't be used "against" them. +1 (as a user) ASF does a hell of a job to discourage participation! I particularly like this gem "the basic rule is that only PMC members have binding votes, *and all others are either discouraged from voting (to keep the noise down)* or else have their votes considered of an indicative or advisory nature only. I'm glad AOO PMC members take these recommendations with a pinch of salt. I particularly dislike the "only" in "advisory nature only". The success of AOO depends on the advice of every community member who cares enough to download a release candidate and test it. - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Pedro Lino wrote: I agree that it is good that some users replicate the building procedure and confirm that it works. ... In any case I couldn't find that particular requirement for a PMC member in the voting page (https://www.apache.org/foundation/voting.html). Maybe that is being too strict? If you read the page Marcus sent, you will basically discover that probably I'm the only one who wrote to the list about all details listed there (such as the NOTICE file etc). Does this mean my vote is "better" than others? Of course not. We don't really look at the status of the voter but at the vote itself and (in case it is a -1) at the reason behind it. So the practice is much saner than the theory and nobody discarded Larry's -1 saying it was not coming from a PMC member. Any correctly justified -1 will of course be treated in the same way, regardless of who casts it. just reading the ASF rules discouraged me to become a committer (I was invited in the past) and still does I did a presentation at ApacheCon a few years ago, called "Bending the Rules: Community over Code over Policy", explaining that policy should never come first and presenting some examples where we had to, say, "interpret" a dumb policy to do the right thing that would not hamper the community. It's still buried at https://community.apache.org/speakers/slides.html and that is still my point of view. But also something that concretely happens in OpenOffice. Agreed. That is why the "*and all others are either discouraged from voting (to keep the noise down)*" is particularly unnecessary (if not offensive). I trust Dave will get that unfortunate wording amended in https://www.apache.org/foundation/voting.html - but again, that is about technical voting (on a code commit) and not release voting. In this sense, the sentence is a bit easier to justify (i.e., if you are expressing an opinion on a controversial code commit you are supposed to have familiarity with the code). Regards, Andrea. - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Hi Pedro, > On Oct 18, 2017, at 1:35 PM, Pedro Lino wrote: > > Hi Dave, all > >> (1) PMC members have a higher bar for their vote. I did not vote at all >> because I did not build from source etc. > > I agree that it is good that some users replicate the building procedure and > confirm that it works. > Replicability is a good measure of transparency and quality. > In any case I couldn't find that particular requirement for a PMC member in > the voting page (https://www.apache.org/foundation/voting.html). Maybe that > is being too strict? > >> (2) Would you name projects where you have felt discouraged? Have you felt >> discouraged here in past years? > > This is the only Apache project I collaborate with (I have been collaborating > with several other FOSS projects for a long time). Actually just reading the > ASF rules discouraged me to become a committer (I was invited in the past) > and still does. > >> We know everyone helping is needed. It was great to have a long standing >> member of the user community like Larry vote to alert us to a problem with >> macOS! > > Agreed. That is why the "*and all others are either discouraged from voting > (to keep the noise down)*" is particularly unnecessary (if not offensive). I just read the page and I agree it is discouraging. I am going to discuss with Community Development about changing the wording. (If that is the correct place.) Regards, Dave > > Regards, > Pedro > > - > To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org > For additional commands, e-mail: dev-h...@openoffice.apache.org > signature.asc Description: Message signed with OpenPGP
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Am 18.10.2017 um 22:35 schrieb Pedro Lino: Hi Dave, all (1) PMC members have a higher bar for their vote. I did not vote at all because I did not build from source etc. I agree that it is good that some users replicate the building procedure and confirm that it works. Replicability is a good measure of transparency and quality. In any case I couldn't find that particular requirement for a PMC member in the voting page (https://www.apache.org/foundation/voting.html). this webpage explains just the general process. Voting for a release is a bit more special and therefore described here: https://www.apache.org/legal/release-policy.html#release-approval Marcus (2) Would you name projects where you have felt discouraged? Have you felt discouraged here in past years? This is the only Apache project I collaborate with (I have been collaborating with several other FOSS projects for a long time). Actually just reading the ASF rules discouraged me to become a committer (I was invited in the past) and still does. We know everyone helping is needed. It was great to have a long standing member of the user community like Larry vote to alert us to a problem with macOS! Agreed. That is why the "*and all others are either discouraged from voting (to keep the noise down)*" is particularly unnecessary (if not offensive). - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Hi Dave, all (1) PMC members have a higher bar for their vote. I did not vote at all because I did not build from source etc. I agree that it is good that some users replicate the building procedure and confirm that it works. Replicability is a good measure of transparency and quality. In any case I couldn't find that particular requirement for a PMC member in the voting page (https://www.apache.org/foundation/voting.html). Maybe that is being too strict? (2) Would you name projects where you have felt discouraged? Have you felt discouraged here in past years? This is the only Apache project I collaborate with (I have been collaborating with several other FOSS projects for a long time). Actually just reading the ASF rules discouraged me to become a committer (I was invited in the past) and still does. We know everyone helping is needed. It was great to have a long standing member of the user community like Larry vote to alert us to a problem with macOS! Agreed. That is why the "*and all others are either discouraged from voting (to keep the noise down)*" is particularly unnecessary (if not offensive). Regards, Pedro - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Hi Pedro, > On Oct 18, 2017, at 12:28 PM, Pedro Lino wrote: > > On 18/10/2017 19:02, Andrea Pescetti wrote: > >> This is OK as long as the Apache Policy doesn't get in the way too much... I >> mean, I highly appreciate when people provide a list of what they did to >> justify their +1. But this shouldn't be used "against" them. > > +1 (as a user) > > ASF does a hell of a job to discourage participation! I particularly like > this gem "the basic rule is that only PMC members have binding votes, *and > all others are either discouraged from voting (to keep the noise down)* or > else have their votes considered of an indicative or advisory nature only. (1) PMC members have a higher bar for their vote. I did not vote at all because I did not build from source etc. (2) Would you name projects where you have felt discouraged? Have you felt discouraged here in past years? > I'm glad AOO PMC members take these recommendations with a pinch of salt. We know everyone helping is needed. It was great to have a long standing member of the user community like Larry vote to alert us to a problem with macOS! Regards, Dave > > Kind regards, > Pedro signature.asc Description: Message signed with OpenPGP
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
On 18/10/2017 19:02, Andrea Pescetti wrote: This is OK as long as the Apache Policy doesn't get in the way too much... I mean, I highly appreciate when people provide a list of what they did to justify their +1. But this shouldn't be used "against" them. +1 (as a user) ASF does a hell of a job to discourage participation! I particularly like this gem "the basic rule is that only PMC members have binding votes, *and all others are either discouraged from voting (to keep the noise down)* or else have their votes considered of an indicative or advisory nature only. I'm glad AOO PMC members take these recommendations with a pinch of salt. Kind regards, Pedro
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
We expect PMC members to have a deep sense of responsibility to their projects. When they cast a binding vote, we trust that they have satisfied themselves enough to warrant them casting the vote. That, alone, should be sufficient, imo. - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
2017-10-18 20:02 GMT+02:00 Andrea Pescetti : > Keith N. McKenna wrote: > >> Even though he is a PMC member his vote cannot be counted as binding as >> he did not build from source as is required by Apache Policy. >> > > This is OK as long as the Apache Policy doesn't get in the way too much... > I mean, I highly appreciate when people provide a list of what they did to > justify their +1. But this shouldn't be used "against" them. > > For example, our last blocker was only affecting binary builds for Mac and > not even all Mac installations; but all of us agreed it was a blocker, even > though it could have gone unnoticed if everybody just did the testing by > Apache Policy. > > So for sure we can keep the tally as it is, but I would tend to just count > the +1s and -1s, and trust that a PMC member will cast a +1 only if he > feels OK with it, rather than basing the tally on the descriptions. In the > end, we really need +1s even from people who only test the binaries. > That's exactly where I stand too. Roberto > > Regards, > Andrea. > > - > To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org > For additional commands, e-mail: dev-h...@openoffice.apache.org > >
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Keith N. McKenna wrote: Even though he is a PMC member his vote cannot be counted as binding as he did not build from source as is required by Apache Policy. This is OK as long as the Apache Policy doesn't get in the way too much... I mean, I highly appreciate when people provide a list of what they did to justify their +1. But this shouldn't be used "against" them. For example, our last blocker was only affecting binary builds for Mac and not even all Mac installations; but all of us agreed it was a blocker, even though it could have gone unnoticed if everybody just did the testing by Apache Policy. So for sure we can keep the tally as it is, but I would tend to just count the +1s and -1s, and trust that a PMC member will cast a +1 only if he feels OK with it, rather than basing the tally on the descriptions. In the end, we really need +1s even from people who only test the binaries. Regards, Andrea. - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
On 10/18/2017 09:29 AM, Keith N. McKenna wrote: On 10/18/2017 11:49 AM, Kay Schenk wrote: Kazunari Hirano (khirano) is on the Apache OpenOffice PMC so his vote should be included in "binding" . MzK "Only the truth will save you now." -- Ensei Tankado, "Digital Fortress" On Oct 17, 2017 18:09, "Carl Marcum" wrote: Voting Results +1: 13 0: 0 -1: 0 +1 (non-binding) Keith N. McKenna (knmc) Kay Schenk (kschenk) Larry I. Gusaas Kazunari Hirano (khirano) Pedro Lino +1 (binding) Jim Jagielski (jim) Patricia Shanahan (pats) Matthias Seidel (mseidel) Mechtilde Stehmann (mechtilde) Marcus Lange (marcus) Andrea Pescetti (pescetti) Carl Marcum (cmarcum) Roberto Galoppini (galoppini) Please let me know if you see any errors or omissions. Thanks to everyone who could participate. Carl - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org Kay; Even though he is a PMC member his vote cannot be counted as binding as he did not build from source as is required by Apache Policy. Regards Keith Ah! OK. Thanks for this notice! Carl's summary is correct. -- -- MzK "Only the truth will save you now." -- Ensei Tankado, "Digital Fortress" signature.asc Description: OpenPGP digital signature
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
On 10/18/2017 11:49 AM, Kay Schenk wrote: > Kazunari Hirano (khirano) is on the Apache OpenOffice PMC so his vote > should be included in "binding" . > > > MzK > > "Only the truth will save you now." > -- Ensei Tankado, "Digital Fortress" > > > > On Oct 17, 2017 18:09, "Carl Marcum" wrote: > > Voting Results > > +1: 13 > 0: 0 > -1: 0 > > > +1 (non-binding) > Keith N. McKenna (knmc) > Kay Schenk (kschenk) > Larry I. Gusaas > Kazunari Hirano (khirano) > Pedro Lino > > +1 (binding) > Jim Jagielski (jim) > Patricia Shanahan (pats) > Matthias Seidel (mseidel) > Mechtilde Stehmann (mechtilde) > Marcus Lange (marcus) > Andrea Pescetti (pescetti) > Carl Marcum (cmarcum) > Roberto Galoppini (galoppini) > > Please let me know if you see any errors or omissions. > > Thanks to everyone who could participate. > > Carl > > > - > To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org > For additional commands, e-mail: dev-h...@openoffice.apache.org > Kay; Even though he is a PMC member his vote cannot be counted as binding as he did not build from source as is required by Apache Policy. Regards Keith signature.asc Description: OpenPGP digital signature
Re: [VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Kazunari Hirano (khirano) is on the Apache OpenOffice PMC so his vote should be included in "binding" . MzK "Only the truth will save you now." -- Ensei Tankado, "Digital Fortress" On Oct 17, 2017 18:09, "Carl Marcum" wrote: Voting Results +1: 13 0: 0 -1: 0 +1 (non-binding) Keith N. McKenna (knmc) Kay Schenk (kschenk) Larry I. Gusaas Kazunari Hirano (khirano) Pedro Lino +1 (binding) Jim Jagielski (jim) Patricia Shanahan (pats) Matthias Seidel (mseidel) Mechtilde Stehmann (mechtilde) Marcus Lange (marcus) Andrea Pescetti (pescetti) Carl Marcum (cmarcum) Roberto Galoppini (galoppini) Please let me know if you see any errors or omissions. Thanks to everyone who could participate. Carl - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
[VOTE][RESULT] Release Apache OpenOffice 4.1.4-RC5 as GA
Voting Results +1: 13 0: 0 -1: 0 +1 (non-binding) Keith N. McKenna (knmc) Kay Schenk (kschenk) Larry I. Gusaas Kazunari Hirano (khirano) Pedro Lino +1 (binding) Jim Jagielski (jim) Patricia Shanahan (pats) Matthias Seidel (mseidel) Mechtilde Stehmann (mechtilde) Marcus Lange (marcus) Andrea Pescetti (pescetti) Carl Marcum (cmarcum) Roberto Galoppini (galoppini) Please let me know if you see any errors or omissions. Thanks to everyone who could participate. Carl - To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org